Abstract:The random forest algorithm (RF) combined with multi-source environmental variables was used to decompose the traditional soil maps of the Qinshui River Basin in Huajiahe Town,Hong’an County,Huanggang City,Hubei Province.The composite soil maps mixed with various soil types were refined,new boundaries were drawn inside the soil polygons to represent a single soil type,and the updated soil map was verified by 373 field sampling points.Results showed that the updated soil map improved its drawing accuracy from 63% to 76%,showing more detailed spatial details and spatial variation information,indicating that using the random forest algorithm to digitally map the traditional soil is feasible and reliable.